GPU Embedded Environmental Impacts data

This repository compiles all data created as part of the alt-impact project (ADEME, CNRS, INRIA) on the environmental impacts of graphics cards (GPUs) used in AI applications. It has been created by TIDE (Etienne Lees-Perasso) and Hubblo.

📕 Read the full report 📕

⚙️ Access the GPU embedded impacts calculator ⚙️

📰 Read the academic article 📰

Repository organisaiton

📁 LCA data/

Contains the raw and processed data used to conduct the Life Cycle Assessment (LCA) of 8 GPUs and 1 GPU rack.

📁 Elementary Analysis/

Contains the results of the elementarty analysis for 3 GPU.

Research partners and acknowledgments

We would like to thank ADEME, CNRS and INRIA, members of the AltImpact program, for the funding and technical support that made this project possible.

We would also like to thank all the stakeholders who contributed to the research project that resulted in the publication of the article on which this work is built : Sophia Falk (University of Bonn), Thibault Pirson (ICTeam), Augustin Wattiez (ICTeam), Lisa Biber-Freudenberger (University of Bonn), Sasha Luccioni (Hugging Face), Aimee van Wynsberghe (University of Bonn), and Julien Comel (TND).

Finally, we would like to thank all the stakeholders in the ecosystem who contributed to the completion of this project. Atos, Boavizta, Bull, Scaleway, Surf, Philippe Desmaison, Aurore Vaudatin (OVH), James Martin (Shippeo), Thomas Brillant (ADEME), Benjamin Paulmier (Carbone 4), Amael Parreaux-Ey (Resilio), Vincent Schellekens (IMEC).

Citation

@misc{lees-perasso2026lcagpu,
        title   = {Analyses de cycle de vie de GPU pour l’intelligence artificielle},
        author  = {Etienne Lees-Perasso, David Ekchajzer, Gauthier Roussilhe, Thomas De Latour},
        year    = {2026},
        publisher = {ADEME},
        url     = {https://github.com/hubblo-org/GPU-Embedded-Environmental-Impacts/}
      }

Licence

This work is licensed under CC BY-SA 4.0